Introduction

Total hip arthroplasty (THA) can significantly reduce pain, improve hip function and, more in general, improve quality of life (QOL) of patients with hip osteoarthritis [1, 2]. Several studies demonstrated such improvement, especially in younger patients [1, 3], as well as the important relationship between QOL, health and hip function [4, 5]. For instance, Reine et al. [6] showed that preoperative activity level could have an impact on postoperative QOL, while Matsunaga-Myogi et al. [1] reported that activity level could increase up to 3 years postoperatively.

QOL in THA was also was usually studied using Short Form 8, Short Form 36, the EuroQol (EQ-5D), or other similar questionnaires [1, 3, 4, 7], while hip function was quantified with tools such as the Harris hip score or the Oxford Hip Score.

Nevertheless, the relationship between spinopelvic alignment and mobility, QOL and clinical outcome of THA is still poorly understood. For instance, results were contradictory on the association between QOL and hip function [8, 9]. This might be due to the fact that hip function scores are external measurements, and they do not take into account the subject’s musculoskeletal conformation. Radiographic-based methods to evaluate spinopelvic mobility [10,11,12], which use sagittal radiographs in different positions (relaxed-seated, flexed-seated, free standing, extension) to measure the lumbar, pelvis and hip mobility, could allow estimating spinopelvic function more accurately, while relating it to pelvic incidence (PI).

Studying preoperative spinopelvic mobility and alignment could also help in predicting postoperative outcome. For instance, Innmann et al. showed that pelvic mobility had an impact on the hip disability and osteoarthritis outcome score one year postoperatively, Ochi et al. [13] reported that preoperative spinopelvic parameters were related to postoperative functional and gait scores, at least in the very short term (3 months).

The aim of this study was to analyze QOL before and after THA, under the hypothesis that radiological pelvic, hip and lumbar mobilities are related to QOL, both preoperatively and postoperatively, and that the analysis of balance and mobility could help to better understand and predict the clinical outcome.

Materials and methods

Patients

This is a prospective and consecutive cohort of patients. Seventy patients with hip arthrosis and with an indication for THA were included between July 2019 and December 2020 at Kyoto City Hospital (Japan). Exclusion criteria were: spinal implant with iliosacral screws, spinal fusion of more than two vertebral levels or scoliosis with coronal Cobb angle higher than 25°. Table 1 reports the demographical data of the cohort. Institutional review board approved the data collection (authorization N. 621).

Table 1 Demographics of the cohort
Full size table

Data collection and radiographic analysis

Full-body lateral radiographs were acquired in free standing position, in extension, and in a flexed-seated position (Fig. 1). For the extension radiograph, patients were asked to hold on to a horizontal bar slightly higher than shoulder level, and they were instructed to extend their pelvis and spine as much as possible. For the radiographs in flexed-seated position, patients were sitting on a stool, and they were instructed to bend forward as far as possible. Acquisitions were obtained preoperatively and six and twelve months after surgery. Status of the contralateral hip was noted (normal, osteoarthritis or THA).

Fig. 1
figure 1

Lateral radiographs in standing position, extension, relaxed seated and flexed seated. Main radiological parameters are reported: pelvic-femoral angle (PFA), T1-pelvic angle (T1PA), lumbar-pelvic angle (LPA), lumbar-femoral angle (LFA)

Full size image

The following standard parameters were measured by an experienced operator in all radiographs: pelvic tilt (PT), pelvic incidence (PI), L1-S1 lumbar lordosis (LL), pelvic incidence minus lumbar lordosis (PI-LL), and T1-pelvic angle (T1PA, Fig. 1) [14]. In addition, the following parameters were computed: femoral sagittal tilt, as the angle between the vertical and a line along the frontal aspect of the first third of the femur diaphysis. The angle was considered positive in flexion and negative in extension [15]. Pelvic-femoral angle (PFA) was also measured, as the angle between the femur and a line drawn from the middle of the sacrum endplate to the centre of the interacetabular hip axis. Finally, lumbar-femoral angle (LFA) was measured as the angle between the femur and a line orthogonal to the L1 upper plateau (Fig. 1). LFA was positive in flexion and negative in extension, which is consistent with LFA being the sum of PFA and the lumbar-pelvic angle (LPA).

Pelvic mobility was considered as “stiff” if the difference between the relaxed-seated and standing SS was lower than 10°, normal between 10° and 30°, and hypermobile when higher than 30° [16]. Pelvic range of motion (ROM) was evaluated as the difference between SS in flexed-seated and extension. Lumbar mobility was considered as “stiff” if the change in LL between standing position and flexed-seated position was lower than 20°, “normal” for a change between 20° and 40°, and “hypermobile” for a change higher than 40° [12]. Hip mobility was defined as the difference in PFA between standing and flexed-seated positions. Hip was considered stiff for mobility < 40°, normal between 40° and 100° and hypermobile above 100°.

Quality of life assessment

Patients were administered the Japanese Orthopaedic Association Hip Disease Assessment Questionnaire (JHEQ)[17] preoperatively and at both postoperative stages. The JHEQ is a QOL assessment method that allows to quantify the patient’s QOL on a scale from 0 to 28 points (increasing with QOL) through 58 questions related to three categories: pain, activity and mental state. A total value can also be computed (with a maximum value of 84). The questionnaire contains general statements which are relevant for the whole population (“Event when I am at rest, my hip is painful”, or “It is difficult for me to climb up and down stairs”) but also a few questions addressing deep flexion and raising from the floor (“It is difficult for me to squat”), which are still relevant for the global population, but specifically targeting the Asian population [19].

Two groups of patients were formed according to their total QOL score change between 6 and 12 months: those that increased their score by more than 10 points and those that decreased by that same amount.

Statistics

Preliminary statistical power analysis suggested that a cohort of 39 patients would allow to detect an improvement of 10 points in total QOL (α = 0.05, β = 0.95)). However, a larger cohort was included to improve the analysis of the correlations between QOL and sagittal alignment.

Differences between QOL parameters at different stages were analyzed with paired Friedman’s test for multiple comparisons. Correlations were quantified with multivariate regression analysis, to correct the confounding effect of age [18]. Stepwise correlation analysis was performed to build a multi-variable model to predict postoperative total QOL from preoperative analysis of sagittal balance.

Significance was set at p < 0.05 and data were reported as median [quartiles]. Calculations were performed with Matlab 2021b (The Mathworks, Natick, USA).

Results

Questionnaires 6 months postoperatively were only available for 62 patients, and 1-year postoperative questionnaires were available for 57 patients (Table 1). Contralateral hip was normal in twenty-nine patients, while twenty showed osteoarthritis and twenty-one had contralateral THA.

Table 2 reports all sagittal parameters measured at each stage and each position. Figure 2 shows QOL preoperative scores, compared to 6 months and 12 months postoperative ones. All QOL items significantly improved 6 months postoperatively (p < 0.0001), and the total score improved in all patients (Fig. 3). Pain was resolved postoperatively for most patients (Fig. 2), while activity and mental status were improved, but they were still not optimal. However, age had to be considered: the activity score was significantly higher in patients younger than 65 (20 [13; 23]) than in older patients (13 [6; 19], p = 0.006). The total score was also slightly higher in younger patients (66 [50; 75] vs 58 [45; 69], p = 0.08), although the difference was not significant.

Table 2 Sagittal parameters measured at each stage (preoperative, 6 and 12 months postoperative) and in three positions: standing (Lat), extension (Ext) and flexed-seated (Flex)
Full size table
Fig. 2
figure 2

Quality-of-life items at three different stages. All quality-of-life items significantly increased six months postoperatively (p < 0.05) and stabilized thereafter; the activity score decreased one year postoperatively, but not significantly. Horizontal dashed lines represent the maximum value for each item

Full size image
Fig. 3
figure 3

Quality-of-life total score (QOL) before and after surgery. Patients with contralateral osteoarthritis or total hip arthroplasty are highlighted as “contralateral issue” (x symbols). The points above the bisector line improved their score; all postoperative scores were higher than the preoperative ones, however, QOL of 40% of patients decreased between 6 and 12 months postoperatively

Full size image

Overall, QOL remained constant between 6 and 12 months (Fig. 2, p > 0.05). However, total QOL decreased by more than 10 points in 16% of patients (N = 8) between 6 and 12 months, while it increased by more than 10 points in 18% (N = 9). Patients for whom the QOL decreased in this period showed lower SS in extension at 6 months than those patients who increased their QOL (20.9 [16.9; 28.0]° vs 32.1 [26.0; 38.6]°, respectively, p = 0.01) and slightly higher SS in flexed-seated position (44.3° [37.8; 50.0]° vs 35.6 [23.5; 51.4]°, p = 0.3). Therefore, patients who decreased their QOL in the longer term showed higher postoperative pelvic range of motion (SSflexion–SSextension = 22.0° [19.9; 24.8]°) than patients increasing their QOL (6.4° [−5.9; 22.2]). No differences were observed in the standing position. Lumbar and hip mobility was not associated to QOL postoperative changes (p > 0.05) (Fig. 4).

Fig. 4
figure 4

Correlations between quality-of-life items (vertical axis) and geometrical parameters (horizontal axis. T1PA T1-pelvis angle; LL lumbar lordosis; PT pelvic tilt; Femur femoral tilt; LFA lumbar-femoral angle). p-values are represented as colours (p < 0.05 in yellow and p < 0.01 in green and p < 0.001 in violet), while correlation coefficients are reports as numbers. Quality of life (QOL) improvement was measured as the difference in score between six months postoperatively and preoperatively. Only parameters with significant correlations are represented

Full size image

Mobility

Pelvic, hip and lumbar mobilities had a minor impact on QOL (Fig. 5): 12 months postoperative, patients with postoperative hypermobile lumbar spine had significantly higher activity scores than normal (p = 0.02) and stiff spines (p = 0.006). Patients with postoperative stiff pelvis had significantly better pain scores 12 months postoperatively than normal pelvis mobility (p = 0.02). A significant correlation was observed between preoperative PT and postoperative change of PT (R = −0.5, p = 0.0001), indicating that patients with low preoperative PT tended to increase their PT postoperative, and vice versa.

Fig. 5
figure 5

Quality of life according to hip, pelvis and lumbar spine mobility. All differences between preop and post-op measurements were significant, and they were not represented in the figure to improve readability. Other significant differences were indicated (* p < 0.05; ** p < 0.01)

Full size image

Correlations

Figure 4 shows the correlations between sagittal alignment parameters and QOL scores. Age was significantly correlated with the postoperative improvement of the total QOL score (r = −0.28, p = 0.08) and with the activity score (R = −0.36, p < 0.003). Stepwise correlation analysis showed that preoperative pelvic tilt alone was the best predictor of postoperative total QOL and of QOL improvement (Fig. 6). However, root mean squared difference between predicted and actual QOL was 14.7.

Fig. 6
figure 6

Correlation between preoperative pelvic tilt and 6 months postoperative total quality of life score and increase of total quality of life (postop–preop). Correlation was adjusted to remove the confounding effect of age, while sex and pelvic incidence did not affect this relationship. Dashed lines represent the 95% prediction interval

Full size image

Discussion

This study analyzed the relationships between QOL, sagittal balance and lumbar spine, hip and pelvis mobility. A previous study showed that QOL, which was quantified using a 8-item. Short Form Health Survey, did not vary between one and three years postoperatively [1]. However, another study by Costa et al. [3] suggested that QOL (measured with EQ-5D) could slightly decrease in the longer term of 5 years. In the present study, sagittal parameters did not significantly change between 6 months and 1 year, but significant differences were observed between preop and 1 year, especially in flexion. This further suggests that postoperative compensation mechanisms continue after 6 months, in particular with an improvement of hip mobility. A possible interpretation is that patients tended to have stiff hip joint before surgery, and hence they had to use their lumbar and pelvic mobilities to compensate. Hip mobility was restored postoperatively, which resulted in a reduction of the number of patients with hypermobile pelvis and spine.

Analysis of correlations confirmed that QOL and sagittal alignment are related; interestingly, it is not the pain or activity scores that showed higher correlation with sagittal parameters, but rather the mental and total scores. Nevertheless, the correlations between the sagittal parameters in flexion and QOL scores confirm that lumbar spine and pelvis mobility plays a significant role in determining the dynamic and functional aspects of QOL. Furthermore, pelvis mobility seems to be relevant in the short-term evolution since patients who decreased their QOL by more than 10 points after 6 months postoperatively also had lower SS in extension and higher SS variation between flexion and extension.

Predicting postoperative QOL remains a challenge. A simple model consisting of preoperative PT is the best predictor of 6-months postoperative QOL and of QOL postoperative change. The prediction might not be sufficiently accurate to provide an expected outcome to the patient, but the uncertainty of the prediction (about 15 points) is similar to the inherent uncertainty of the JHEQ test (also 15 points, [20]). In general, patients can expect a better QOL improvement if they present low PT, T1PA and PI-LL angles, as well as low PT in extension.

This study presents some limitations. Firstly, postoperative QOL questionnaires were not available at both postoperative time points for all patients. Secondly, patient reported outcome questionnaire suffers from inherent low reproducibility: the bias of test–retest JHEQ questionnaire was previously estimated to be 1.5 points, with an uncertainty of about ± 15 points [20]. Nevertheless, clear trends were observed in the postoperative improvement of QOL, as well as in the relationship between QOL and radiological parameters. Finally, this study was conducted on a relatively small cohort, and it was monocentric, which could limit the generalizability of the results.

Conclusion

This study confirmed that QOL and spinopelvic alignment and function are inter-related in THA. All patients improved their QOL after THA, but results suggest that patients with low preoperative PT, T1PA and PI-LL might benefit from an increased QOL improvement. Results also suggest that QOL can decrease between 6 and 12 months postoperatively in 16% of patients, while other patients continued improving in this time. Further studies could elucidate the functional determinants of this QOL deterioration, albeit small, in the medium to long term.